GNOME Bugzilla – Bug 580198
Crashes when going online after offline message movement in IMAP
Last modified: 2009-06-09 10:59:44 UTC
What were you doing when the application crashed? Trying to reproduce Fedora bug #497598 after upgrading to rawhide. Yes, it's still there, but now with a nice GUI. Distribution: Fedora release 10.92 (Rawhide) Gnome Release: 2.26.1 2009-04-13 (Red Hat, Inc) BugBuddy Version: 2.26.0 System: Linux 2.6.30-rc3-wl #93 SMP Sat Apr 25 02:58:42 EDT 2009 x86_64 X Vendor: The X.Org Foundation X Vendor Release: 10503000 Selinux: No Accessibility: Disabled GTK+ Theme: Nodoka Icon Theme: Fedora GTK+ Modules: canberra-gtk-module, pk-gtk-module, gnomebreakpad Memory status: size: 1181769728 vsize: 1181769728 resident: 28979200 share: 19755008 rss: 28979200 rss_rlim: 18446744073709551615 CPU usage: start_time: 1240645070 rtime: 54 utime: 40 stime: 14 cutime:0 cstime: 2 timeout: 0 it_real_value: 0 frequency: 100 Backtrace was generated from '/usr/bin/evolution' [?1034h[Thread debugging using libthread_db enabled] [New Thread 0x7fceecb57910 (LWP 2603)] [New Thread 0x7fcefd0d9910 (LWP 2598)] [New Thread 0x7fceee5f1910 (LWP 2595)] [New Thread 0x7fceeedf2910 (LWP 2594)] [New Thread 0x7fceef5f3910 (LWP 2593)] [New Thread 0x7fceefdf4910 (LWP 2592)] [New Thread 0x7fcefdb02910 (LWP 2589)] [New Thread 0x7fcefe303910 (LWP 2588)] 0x00007fcf0b3ecca3 in poll () from /lib64/libc.so.6
+ Trace 214778
Thread 2 (Thread 0x7fceecb57910 (LWP 2603))
----------- .xsession-errors (304 sec old) --------------------- XIO: fatal IO error 11 (Resource temporarily unavailable) on X server ":0.0" xterm: fatal IO error 11 (Resource temporarily unavailable) or KillClient on X server ":0.0" XIO: fatal IO error 11 (Resource temporarily unavailable) on X server ":0.0" XIO: fatal IO error 11 (Resource temporarily unavailable) on X server ":0.0" after 144 requests (144 known processed) with 0 events remaining. after 247 requests (247 known processed) with 3 events remaining. after 5566140 requests (5566140 known processed) with 55 events remaining. The application 'gnome-screensaver' lost its connection to the display :0.0; most likely the X server was shut down or you killed/destroyed the application. icewm-session: Warning: File still open: fd=4, target='pipe:[6973]' (missing FD_CLOEXEC?) icewm-session: Warning: Closing file descriptor: 4 icewm-session: Warning: File still open: fd=5, target='pipe:[6973]' (missing FD_CLOEXEC?) icewm-session: Warning: Closing file descriptor: 5 icewmbg: Can't open display: :0. X must be running and $DISPLAY set. --------------------------------------------------
More information at https://bugzilla.redhat.com/show_bug.cgi?id=497598
Created attachment 135096 [details] [review] proposed eds patch for evolution-data-server; Highlights: a) folder's lock on refresh_info added. it's required because IMAP's journal replay uses camel_folder_get_message, which locks this lock. There can happen that threads will hold folder's lock, but will wait connect_lock, but the other has this locks opposite. Thus adding lock there. b) shouldn't replay journal always c) busy-lock, when journal's message missing in summary, fixed d) get_matching translates from local summary UIDs to real UIDs too e) journal_decode_folder uses its 'folder' when asked for it, doesn't open new for the operation Side-effect: IMAP changes done in offline should work much better than before.
Milan, seems awesome right. I have a hack in camel-local-folder.c:local_refresh_info where I lock it there. Just remove the lock. It was originally in camel-folder/refresh_info but there was a ugly hang, when replaying with disco. Since we moved away from disco to offline, it should have fixed it. So remove that hack and commit to master.
Created commit 7aaf3c8 in eds master (2.27.3+) Remove the code you asked to remove from camel-local-folder.c:local_refresh_info.